RESEARCH Open Access

A study on the embarrassment of senderswho missend emojis with oppositemeanings on social apps: taking WeChat asan exampleLiyuan Liu1,2 , Yen Hsu2*, Jie Zhang3 and Qianling Jiang4

Abstract

With the increasing popularity of social apps, sending emojis has become a very common way of expressing one’semotions. However, situations often arise when people send the wrong emoji by mistake, or sometimes even anemoji with an opposite meaning, which can cause embarrassment to the sender. Taking WeChat as an example,which is widely used in Chinese communities, this study summarizes 10 types of dialogue situations in which themeaning of an emoji is wrongly sent and 12 types of emotional components that are related to embarrassment.The purpose of this study was to analyze the extent to which the 12 emotional components that are associatedwith embarrassment actually explain what embarrassment is, as well as the different degrees of embarrassmentamong the different genders and age groups. The results showed that (1) among the emotional components ofembarrassment, shame has the highest explanation degree for embarrassment; (2) males are more likely to beaffected by embarrassment than females; and (3) users aged 18–25 and 26–30 years are more likely to be affectedby embarrassment than those aged between 31 and 40 when they mistakenly send WeChat emojis. This studyprovides a reference value for their sustainable psychological impact on social app users.

Keywords: WeChat, Emoji with opposite meaning, Missending, Sender, Embarrassment, Interaction design

IntroductionAmidst the fierce competition of mobile real-time socialsoftware, it is becoming more important to maintain theusers’ willingness to continue using the software (Zhanget al., 2017). Mobile Instant Messengers (MIMs), such asMessenger, WhatsApp, LINE, and WeChat, are very popu-lar, and sending emojis has become a very common way ofexpressing one’s emotions (Annamalai & Salam, 2017). Re-searchers have found that while emojis are very popular,the wrong ones or even those with an opposite meaningare sometimes mistakenly sent by users, when one is notexpecting to do so; it may cause embarrassing emotions.

Contact and communication with others are importantactivities in the lives of people, and information transmis-sion can be facilitated by means of WhatsApp, WeChat,LINE, and Telegram. In China, WeChat has dramaticallychanged peoples’ social lifestyles, and contact has becomean indispensable part of their daily lives (Danesi, 2016).According to data from Tencent in March 2018, WeChathas more than 1 billion active users worldwide every day.WeChat is the most important mobile MIMs in theChinese-speaking world (Wang et al., 2019). According to asurvey, around 25% of the users open WeChat over 30times per day, and 55.2% use WeChat over 10 times perday (Wen et al., 2016). However, most past studies haveonly focused on Facebook, YouTube, or other internationalMIMs that are most popular in Western countries (Montaget al., 2018; Tsai & Men, 2018). Few studies have focused

© The Author(s). 2020, corrected publication 2020. Open Access This article is licensed under a Creative Commons Attribution4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, aslong as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence,and indicate if changes were made. The images or other third party material in this article are included in the article's CreativeCommons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's CreativeCommons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will needto obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

* Correspondence: yhsu@gm.ttu.edu.tw2The Graduate Institute of Design Science, Tatung University, Taipei, TaiwanFull list of author information is available at the end of the article

Psicologia: Reflexão e CríticaLiu et al. Psicologia: Reflexão e Crítica (2020) 33:20 https://doi.org/10.1186/s41155-020-00159-4

on the beneficial effects of Chinese MIMs on its users, es-pecially WeChat (Research, 2016).Pang (2018a, 2018b) showed that there are four factors

that describe a person’s willingness to use WeChat,namely, time, emotion, social interaction, and fashion.Research shows that 69.4% of users use WeChat daily;12.2% use it three to four times a week; and 11.6% use itfive to six times a week. Although the frequency of theuse of WeChat per week can be better predicted bychanges in the time spent (e.g., whether it was for pleas-ant rest, leisure, and relaxation, or for helping andthanking others) (Zhang et al., 2017), the main motiv-ation of 84.4% of the users was to have an emotionalconnection with everyone they knew (Hossain et al.,2019). The interaction between complex emotions andambiguous information comprehension makes the infor-mation more ambiguous and further affects the cogni-tion of emojis (Zhang et al., 2017).It is very common and popular to communicate by

using emojis. From smartphones to social media, smalldigital graphics are ubiquitous and people are very willingto send these graphics (Pang, 2018a). The study showedthat 90% of the people found communicating with emojismore meaningful (Pang, 2016). Emojis not only serve as auser’s emotional guide, but also provide many other com-municative functions (Riordan, 2017a). Although there areenough graphics available for downloading, each graphicmight not be used for a long time, and some downloadedemojis will be discarded, but built-in emojis will be keptand continue to be used (Tseng & Hsieh, 2019).The symbol for a smiling face as an emoticon is “^_^”,

while as an emoji, it is a small symbol represented by apicture, such as the symbol for a smiling face “ ”. Emojisare used in almost the same way as emoticons (Jaeger et al.,2018), but it does not mean that they are used with thesame frequency, or that they have the same influence. Pre-vious studies have shown that Twitter users using emojishave reduced their use of emoticons because emojis are re-placing emoticons in fulfilling the same paralinguistic func-tions (Danesi, 2016). It has been noted that the largenumber of emojis that describe a wide range of content hasreduced the need for the use of emoticons (Pavalanathan &Eisenstein, 2016) and that emojis are now more commonlyused than emoticons (68.1% vs 30.9%) (Li & Yang, 2018).Although digital dialogues are mainly by means of writtencommunication, non-verbal communication, such as viaemojis, can facilitate emotional expression and improve theefficiency of dialogues (Chen et al., 2017), and 11.3% ofpeople use emojis to shorten the time it takes to send amessage (Ganster et al., 2012).This study conducted psychological research on the emo-

tional communication function of WeChat emojis, for thefollowing reasons: (1) emojis are used by almost all socialmedia users to start and maintain conversations and to

facilitate interpersonal relationships (Wang et al., 2019); (2)during network-based communications, the use of emojisexhibits a high frequency, function, and efficiency (Choiet al., 2017; Hanna et al., 2017; Lomanowska & Guitton,2016; Yang & Lee, 2020); (3) the communication functionof emojis, therefore, enhances the emotional relationshipbetween people (Dhir & Tsai, 2017; Pang, 2018b); and (4)during emotional communication in the social environ-ment, people often use positive emojis to create a positiveatmosphere and to promote harmonious relations (Kelly &Watts, 2015). However, misunderstanding the meaning ofan emoji may cause communication barriers, and in specificsituations, it may damage relationships (Li & Yang, 2018).As there are only a few research studies on the psycho-logical effects of WeChat on users (Danesi, 2016; Kayeet al., 2016), this paper looks to fill a gap in the related lit-erature through an analysis of missent emojis.Previous studies have emphasized the functional role,

the condition for use, and the usefulness of emojis in fa-cilitating communication (Li & Yang, 2018). However, fewstudies have looked at the impact on the emotions of thesender of missending WeChat emojis with an oppositemeaning. The emotional role of an emoji is expounded insociological theory, which notes that the time and effortspent on the use of emojis can help to maintain andstrengthen social relations (Tigwell & Flatla, 2016).Embarrassment is an important emotion that affects social

communication, and it plays a very important role in main-taining and developing social relations (Pang, 2018a). Re-search has found that embarrassment will lead to negativeconsequences (Castillo et al., 2013; Derks et al., 2007; Pau-leen & Yoong, 2001; Tseng & Hsieh, 2019). Previous re-search by Riordan (2017b) has pointed out thatembarrassment is a combination of such emotions as bewil-derment, clumsiness, shame, and annoyance. The negativeconsequences of embarrassment will lead to a decline in theusers’ satisfaction with WeChat, and this will have an impacton their willingness to continue using WeChat (Bastin et al.,2016). WeChat users have complained on social media, thatthey have sent wrong emojis because the WeChat emojisthat have opposite meanings are arranged close to eachother. The contribution of this study is that it not only hastheoretical significance in network social psychology, but italso has a practical significance for improving the WeChatusers’ willingness to continue using WeChat.However, we know little about the differences be-

tween the emoji usage of men and women (McCam-bridge & Consedine, 2014) and that different emojisare used by different ages (Miller, 1996). In order tomake up for the lack of studies on the differences inemoji usage among users of different genders andages and to improve the continuous willingness ofusers to use WeChat emoji, the main purposes of thisstudy are as follows:

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 2 of 15

1. To explore which gender is more easily affected byembarrassment after missending a WeChat emojiwith an opposite meaning;

2. To explore which age group is more easily affectedby embarrassment after missending a WeChatemoji with an opposite meaning; and

3. To analyze the senders’ embarrassment aftermissending WeChat emojis, in order to determine theemotional components that are related toembarrassment and their degree of influence.

This study can help to promote people’s willingness tocontinue using WeChat and to provide a theoretical basisfor developing a good social networking situation. Thefindings of this study will provide a reference for promot-ing positive emotions in future MIMs.

MethodsParticipantsThe Internal Review Board of the Neural Ergonomics La-boratory has approved this study. This study adopts theonline questionnaire survey format. All participants clickthe webpage link of the survey to view the survey descrip-tion of the study; they voluntarily answer the researchquestions and can quit the survey at any time. Therefore,all participants agreed to participate in the study underthe principle of fully informed and voluntary participation.To study the link between gender differences and embarrass-

ment, this study treated 30 men and 30 women as the subjects.Our study surveyed participants in 10 conversation

scenarios involving relationships between family mem-bers, relatives, lovers, friends, and leaders. In order tomeet the requirements of the sample in this study, it isnecessary for participants to master all the interpersonalrelationships in these 10 dialogues. Therefore, from the378 samples, we selected a number of samples that metthe research criteria and which includes all the interper-sonal experiences in 10 conversation scenarios.After this step, our selected sample covered interpersonal

relationships in 10 conversation scenarios, which met theneeds of the study. However, not every participant met thisrequirement. Most participants had difficulty matching theinterpersonal relationships in all the conversation scenarios.Of the 378 samples, only a few matched the interpersonalrelationships covered by the 10 conversation scenarios.In a sample of 378 questionnaires, we first selected a sam-

ple of 30 women who exactly matched the interpersonal re-lationships contained in the 10 conversation scenarios. Inpsychological research, the common practice of defining asample is usually to include the same number of men andwomen in a single sample (Sigel, 1965). In order to keep thesample size of men and women equal, we conducted a sec-ond round of questionnaire screening, by selecting a sampleof 30 men, which exactly matched the interpersonal

relationships contained in the 10 conversation scenarios.Therefore, the final sample of this study was 30 men and 30women, which was consistent with the number of sampleplans in this study.

Procedure and materialsThis study is divided into four stages, namely (1) the emojilayout of WeChat and the determination of their oppositemeanings; (2) the collection and analysis of dialogue scenar-ios that were caused by mistakenly sending WeChat emojiswith opposite meanings; (3) the identification of the emo-tional components related to embarrassment; and (4) thepreparation of questionnaires and the respondents answer-ing the questions by selecting the intensity of the emotionalcomponents, according to the dialogue scenarios.

Phase 1—the emoji layout of WeChat and thedetermination of its opposite meaningThe dialogue interface for sending a WeChat emoji isshown in Fig. 1. The area that is surrounded by the red dot-ted line is the emoji group. Figure 2 shows the layout of allthe emojis of WeChat. There are 104 built-in basic emojisin WeChat, which are divided into five groups (Fig. 2), andthey can be selected by sliding the cursor left and right.Groups One, Two, Three, and Four have three lines each,with 23 emojis in each group. The last position of the lastline of each group (the lower right-hand corner of the emojigroup) is occupied by the deletion symbol. Group Five onlyhas one line and is comprised of eight emojis.Based on the perspective of the emoji sender, this study de-

termined the opposite meaning of an emoji from a specificgraphic meaning, context, and culture. (1) In WeChat, everyemoji has a specific meaning. Pressing the emoji for a longtime will prompt the meaning of the emoji. Users send theemojis based on what they mean, such as “ ” for strongand “ ” for weak. By knowing the specific meanings ofemojis, users can clearly determine their opposite meanings,without ambiguity and polysemy. (2) When actually usingthe emojis, their meanings are very contextual, sometimeswith complex, multi-layered or overlapping meanings (Darics,2013). For example, “ ” can mean not only surprise andconfusion, but also doubt (Tannen, 1994). Therefore, thecontext is very important for understanding emoticons anddetermining their opposite meanings. (3) Emojis have differ-ent meanings in different cultures (Ljubešić & Fišer, 2016).Emoticons reveal cultural diversity (Freedman, 2018). For ex-ample, “ ” has five main meanings in the world, with“good” being the dominant interpretation (Morris, 1979), butthe “thumbs up” in Iraq or in Sweden is an insult and means“up yours”. The respondents in this study are ChineseWeChat users who have a common cultural background, sothey have a relatively consistent understanding of the oppos-ite meanings of WeChat emojis.

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 3 of 15

Emojis with opposite meanings were collected from all dia-logue situations, including two types, namely (1) the dialoguesituation captured in WeChat screenshots and (2) the dia-logue situation described by the sender through a text. Ac-cording to the specific meaning, context, and culture of

emojis, we analyzed their opposite meanings in the dialoguesituations that were collected. We extracted two emojis withopposite meanings in each dialogue situation and took thetwo emojis with opposite meanings in each situation as a pair.Through the analysis of the layout position of emojis with op-posite meanings in the dialogue situation, we found that indifferent contexts, users mistakenly sent emojis with oppositemeanings. This occurred mainly between two adjacent emo-jis, including the upper and lower adjacent emojis and the leftand right adjacent emojis. As Groups One, Two, and Threehave three lines each, there may be cases of missending twohorizontally or vertically adjacent emojis, and because GroupFive has only one line, the case of missending vertically adja-cent emojis does not occur.There are two kinds of emojis with opposite meanings

that may be mistakenly sent by users. The one is when thetwo adjacent emojis have obviously opposite meanings invisual form, i.e., they are intuitive and easily identifiableemoji graphics, including four pairs, and the other is thatthe opposite meanings of two emoji graphics are not intui-tive in visual form, but will cause opposite results accordingto the dialogue situation (Tseng & Hsieh, 2019). There are25 pairs. In summary, there are 29 pairs of emojis with op-posite meanings in WeChat, as can be seen in Fig. 3.

Phase 2—the selection and analysis of situationsThe collection of dialogue scenarios on embarrassingtopics is very sensitive as it involves interpersonal

Emoji group

Fig. 1 The WeChat (Version: 6.7.4) interface and its field of emojis

Group 1 Group 2

Group 3 Group 4

Group 5

Fig. 2 WeChat’s built-in basic emoji groups

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 4 of 15

relationships and personal emotions (Bethell et al., 2014;Uysal et al., 2014). There is sufficient evidence to show thatthe transmission of information is usually related to emo-tions and identities (Bas-Hoogendam et al., 2018; Hofmannet al., 2006; Sharkey & Singelis, 1995). In this study, theanonymity of the interviewees is the key factor, especiallyregarding the integrity and accuracy of the collection of dia-logue scenarios. Therefore, the dialogue scenarios are col-lected by anonymous online questionnaires. There are twomain forms of respondent feedback on the dialogue scenar-ios, namely (1) the respondents who described, in writing,about the dialogues in which they had sent WeChat emojiswith opposite meanings by mistake; and (2) the respon-dents took screenshots of the dialogues in which they hadsent WeChat emojis with opposite meanings by mistake. Intotal, 103 anonymous online questionnaires were collectedwith conversational scenarios that involved the missendingof WeChat emojis with opposite meanings.In the 103 conversations, the more frequently a pair of

emojis with opposite meanings appeared indicated that thepair was more likely to be confused by the sender. In theconversation situations that we collected, the opposite emojiswith “ ” appeared the most, 14 times in total, with arelative frequency of 13.59%. Compared with other emojiswith opposite meanings, emojis are the most likely to be con-fused by the sender. Emojis “ ”, “ ”, “ ” and“ ” also appear with relatively high frequency, suggestingthat these emojis are also easily confused by the sender.Based on the 103 collected conversations, we studied

emojis and contexts as a whole and took the seven inten-tions for sending emojis (Hu et al., 2017) as a referenceindex to analyze the different psychological states and fre-quency of sending emojis with opposite meanings. The re-sults are shown in Table 1. Emojis with opposite meanings

have the highest frequency of expressing emotions(34.95%), while those with opposite meanings have a higherfrequency of strengthening the expression (22%). The top10 most-used emojis are mostly positive, while the bottom10 are mostly negative. The smiley emoji “ ” was the mostused (63%) (Lin, 2019).We also measured how often emojis with opposite

meanings appeared in different relationships. An analysisof 103 conversation scenarios found that the interper-sonal relationships in which the expressions with oppos-ite meanings were incorrect, included friends, relatives,colleagues, lovers, and leaders, which are common innetwork communication (Hongqiang et al., 2019).As shown in Table 2, the frequency of sending emojis

with opposite meanings is the highest among friends(30.10%). The second highest is the relative frequency be-tween relatives (18.45%), and the lowest frequency was forthe leadership category, with a relative frequency of 8.74%.This could be explained by the fact that emojis are usedmore in intimate relationships and less when communicat-ing with people who are in a distant relationship. In task-oriented formal communication, in order to avoid commu-nication difficulties caused by the different interpretation ofemojis, people may prefer to use meaningful words toclearly convey the information (Hongqiang et al., 2019).The location of the WeChat emojis in context is divided

into four categories, namely, the single, beginning, middle,and end. WeChat emojis are rarely seen in the beginningand the middle. Emojis were not used in combination withother words, and the relative frequency with which they ap-peared separately was 34.2%, while the relative frequencywith which they appeared at the end was the highest(51.7%) (Hongqiang et al., 2019). This is consistent with theposition of emojis that were collected in the dialogue scenes

Fig. 3 Classification of WeChat emojis with opposite meanings

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 5 of 15

of this study, while the positions of emojis with oppositemeanings sent by the sender appeared mainly at the end ofthe dialogue.Three linguistic experts were invited as judges to analyze

and classify the themes of the 103 conversations, accordingto the relationship between the sender and receiver and tosummarize the dialogue scenarios with a similar content.The top 10 situations with the most similar topics and thosethat occurred most frequently were selected (see Appendix).The 10 dialogue scenarios contained common interpersonalrelationships and were representative in dialogues where themissending of emojis with opposite meanings existed. Forexample, the user originally wanted to send “ ”, indicatinga compliment, but actually mistakenly sent “ ”, indicatingthe opposite meaning. Table 3 lists three dialogue scenariosto exemplify the occurrence of embarrassment.In order to protect the privacy of the interlocutors and to

avoid the influence of the personal profile photos, names, andnicknames of the respondents, their personal information,such as profile photos, names, or nicknames, in the dialoguescenarios were reprocessed. The content and explanations ofthe 10 dialogue scenarios are listed in the Appendix below.

Phase3—determination of the emotional componentsA consideration concerns the construct that psychologicaltests purport to measure (Miguel & Pessotto, 2016).Through a literature review and analysis, this study

determined the main sources of the following 12 emotionalcomponents that are related to embarrassment: (1) the fol-lowing eight embarrassment-related emotional componentsare proposed by Grace (2007), namely, “Angry”, “Humili-ated”, “Upset”, “Self-conscious”, “Foolish”, “Frustrated”,“Depressed”, and “Shocked” and (2) the four emotionalcomponents related to embarrassment are proposed byother literature, namely, “Distressed”, “Fearful”, “Anxious”,and “Ashamed”. These are shown in Table 4.

Phase 4—questionnaire design and surveyIn order to avoid the situation where participants focus onlyon answering the first few questions and then possiblyweaken their answers to the subsequent questions aboutembarrassing emotions, the questionnaire randomly ar-ranged the words for the 12 emotional components of em-barrassment, or that are related to embarrassment, toensure the quality of the questionnaire responses. Accordingto the degree, from strong to weak, the respondents chosethe strength of each word to describe the situation. Thequestionnaire of this study adopted the Likert 7-point scaleto divide the level of strength of the emotional componentsthat are related to embarrassment. The questionnaire datawere collected in an anonymous manner by a professionalonline research company (https://www.sojump.com/),which has rich experience in working in China (Borg et al.,1988; Higuchi & Fukada, 2002; Kaufman, 2004; Lewis, 1971;Modigliani, 1971). By using the network’s questionnaire sur-vey, we actually collected a total of 378 samples.

Statistical analysisThe total reliability coefficient of the data is 0.945, which in-dicates that the reliability of the research data is very high.For the research data, we conducted the following: (1) a re-

gression analysis to analyze the impact of 12 embarrassment-related emotional components on embarrassment and (2) atwo-way ANOVA to analyze the relationship between inde-pendent variables (gender and age) and embarrassment.

ResultsThrough experiments and analysis, it was found that (1)the average value of embarrassment and its emotionalcomponents for males is significantly higher than that forfemales; (2) compared to users between 31 and 40 yearsold, users aged between 18 and 5 years are more suscep-tible to embarrassment when they mistakenly sendWeChat emojis, while users aged between 26 and 30 yearsare also susceptible to embarrassment; and (3) the degreeof shame caused by embarrassment is the highest.

Gender and embarrassmentFrom the questionnaires on the relationship between gen-der and embarrassment components, a statistical analysiswas carried out on the questionnaires of the 30 male and

Table 1 Psychological state of the sender when sending emojis

Frequency Relative frequency

Emotional expression 36 34.95%

Strengthen the expression 22 21.36%

Adjust the mood 10 9.71%

Express a sense of humor 5 4.85%

Express satire 0 0.00%

Express intimacy 8 7.77%

Describing the content 9 8.74%

Others and unknown 13 12.62%

Total 103 100%

Table 2 The frequency of emojis with opposite meanings indifferent relationships

Relationships Frequency Relative frequency

Friends 31 30.10%

Relatives 19 18.45%

Colleagues 15 14.56%

Lovers 10 9.71%

Leaders 9 8.74%

Others and unknowns 17 16.50%

Total 103 100%

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 6 of 15

female participants, respectively. The analysis was presentedvisually by using data analysis technology by Python soft-ware, as shown in Fig. 4.The explanation of Fig. 4 is as follows: the X-axis shows

the 10 dialogue scenarios that were used to test embarrass-ment; the Y-axis is embarrassment and its emotional com-ponents; the color blocks represent the P values, rangingfrom 0 to 1, from white to blue. The two-way ANOVA wasused to study the different gender aspects in 1 + 12 items,namely, Embarrassed, Shocked, Frustrated, Depressed,Foolish, Self-conscious, Upset, Humiliated, Angry,Ashamed, Anxious, Fearful, and Distressed. From the two-way ANOVA thermogram, the following can be seen: inScenario 6, gender is significant in Humiliated, which indi-cates that the subjects of different genders exhibit differ-ences in being Humiliated; in Scenario 8, gender issignificant in Foolish, which indicates that the subjects ofdifferent genders show differences in feeling Foolish; exceptfor the above differences, different genders do not show

differences in feeling Embarrassed, Shocked, Frustrated,Depressed, Foolish, Self-conscious, Upset, Humiliated,Angry, Ashamed, Anxious, Fearful, and Distressed (P >0.05), which indicates that subjects of different genders arerelatively consistent and have no differences in the aboveemotions.Although the smaller the P value, the more significant

the statistical test is, it is difficult to quantify the P value inthe probability scale as the data strength for the nihilityhypothesis. Even if the null hypothesis is rejected, there isstill a high probability (about 20%) that null is true (Grace,2007). Therefore, we adopted the mean values of femaleand male for different emotions under 10 dialogue scenar-ios, graphed them in Fig. 5 by using data analysis technol-ogy and the function of data visualization by Python, andthen conducted a further analysis.The explanation of Fig. 5 is as follows: the X-axis repre-

sents embarrassment and its emotional components; andthe numbers 2, 3, 4, 5, and 6 on the Y-axis represent thescale scores. When the average score of males is higherthan that of females in the same scenario, the red solid lineis used to show a connection; a larger red circle is used torepresent the score value of the males; when the averagescore of the females is higher than that of males in the samescenario, the blue solid line is used for a connection; andthe larger blue circle is used to represent the score value ofthe females. Figure 5 shows that there are obviously morered lines than there are blue lines. Therefore, the averagevalue of embarrassment and emotional components ofmales is significantly higher than that of the females.

Age and embarrassmentThe explanation of Fig. 6 is as mentioned in the previoussection. The two-way ANOVA was used to study the differ-ences of 13 items, including Shocked, Frustrated, De-pressed, Foolish, Self-conscious, Upset, Humiliated, Angry,Ashamed, Anxious, Fearful, Distressed, and Embarrassed.The two-way ANOVA thermogram shows the following: inScenario 10, different ages show significance in Anxiousand Fearful, which indicates that the subjects of different

Table 3 Examples of dialogue scenarios involving the missending of WeChat emojis

No. Right Opposite(Missent)

Example

1 The mother told the father that “our son got first place in the competition today”. The father wanted to send toshow his compliment to the son’s performance, but mistakenly sent with the opposite meaning, resulting inembarrassment.

2 A couple was sending messages to each other. The girl texted, “Keep warm. Miss you”. The boyfriend wanted tosend “OK, I love you ”, but mistakenly sent “OK, I love you ”, which led to the embarrassment of the boyfriend.

3 A friend asked “Are you free tonight? Welcome to my birthday party.” The receiver wanted to reply with the graphicof a birthday cake to express the best wishes for the birthday, but mistakenly sent “ ”, resulting inembarrassment.

Table 4 Twelve embarrassment-related emotional components

No. Emotional componentsof embarrassment

References

1 Angry (Bethell et al., 2014)

2 Humiliated (Bethell et al., 2014; Uysal et al., 2014)

3 Upset (Bastin et al., 2016; Higuchi & Fukada,2002; Kaufman, 2004; Lewis, 1971;Modigliani, 1971)

4 Self-conscious (Grace, 2007)

5 Foolish (Grace, 2007)

6 Frustrated (Babcock, 1988; Grace, 2007; Miller, 1996;Parrott & Smith, 1991)

7 Depressed (Grace, 2007)

8 Shocked (Grace, 2007)

9 Distressed (Babcock, 1988; Grace, 2007; Miller, 1996;Parrott & Smith, 1991)

10 Fearful (Grace, 2007)

11 Anxious (Grace, 2007)

12 Shame (Bethell et al., 2014)

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 7 of 15

ages exhibit differences in Anxiety and Fearfulness; exceptfor the above differences, subjects of different ages do notshow any significance in Shocked, Frustrated, Depressed,Foolish, Self-conscious, Upset, Humiliated, Angry, Shame,Anxious, Fearful, Distressed, and Embarrassment, which in-dicates that the subjects of different ages are consistent andthat there are no significant differences in Embarrassment,Shocked, Frustrated, Depressed, Foolish, Self-conscious,

Upset, Humiliated, Angry, Shame, Anxious, Fearful, andDistressed.Combining the description about the P value in the previ-

ous section, we adopted the mean value of the respondentswith the three age groups for different emotions under 10dialogue scenarios. Figure 7 is plotted by using data analysistechnology and the function of data visualization by Pythonsoftware and conducted further analysis.

Fig. 4 The two-way ANOVA thermogram of gender and embarrassment components

Fig. 5 Mean values of females and males for different emotions under 10 dialogue scenarios

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 8 of 15

The explanation of Fig. 7 is as follows: the X-axis repre-sents embarrassment and its emotional components; the Y-axis represents the scale scores; when the average score ofan age group is higher than that of the other two age groupsin the same scenario, it is represented by a larger circle. The31–40-year age group has small green circles in all scenarios,indicating that it has lower scores, and the average scores forEmbarrassment and emotional components are obviouslylower than those for the 18–25-year-old and 26–30-year-oldage groups. These age groups are basically consistent inshowing larger blue and red circles in each scenario, whichindicates that the emotional evaluations of both groups areconsistent. In summary, the average score of age groups inEmbarrassment or certain emotional components is “18–25-years olds ≈ 26–30-year olds > 31–40-year olds”.

Regression analysisAs shown in Fig. 8, the regression analysis was made by tak-ing Shame, Depressed, Anxious, Upset, Shocked, Foolish,Humiliated, Self-conscious, Frustrated, Fearful, Distressed,and Angry as the independent variables and Embarrass-ment as the dependent variable. Among them, the degreeof explanation of Shame for Embarrassment is 0.87, that ofDepressed for Embarrassment is 0.86, that of Anxious andUpset for Embarrassment is 0.85, that of Shocked for Em-barrassment is 0.84, that of Foolish for Embarrassment is0.83, and that of Humiliated for Embarrassment is 0.82,while the degrees of explanation of Self-conscious, Frus-trated, Fearful, and Distressed for Embarrassment are be-tween 0.78 and 0.71, at 0.78, 0.74, 0.73, and 0.71,

respectively. The degree of explanation of Angry for Em-barrassment is 0.58, which is the lowest and thus has acomparatively lower explanation for Embarrassment, com-pared with other emotions.This study presented a thermodynamic diagram of correl-

ation coefficients by using data analysis technology and thefunction of data visualization by Python software, as shownin Fig. 9. It examined 12 emotions, namely, Angry, Dis-tressed, Fearful, Frustrated, Self-conscious, Humiliated,Foolish, Shocked, Upset, Anxious, Depressed, and Shame.The blocks of different colors indicate the strength of thecorrelation; red indicates the high correlation between thetwo emotions; blue indicates the low correlation betweenthe two emotions; and yellow indicates the moderate correl-ation between the two emotions. As can be seen in Fig. 9,most of the color blocks are red, which implies that the cor-relation is generally high. The 11 emotions of Angry, Dis-tressed, Fearful, Frustrated, Humiliated, Foolish, Shocked,Upset, Anxious, Depressed, and Shame have a correlationcoefficient of more than 0.7, which shows a high correlation.The correlation coefficients of Self-conscious and otheremotions are generally low, of which the lowest correlationcoefficient is for Distressed, which is 0.39 and marked indark blue, followed by Angry, Frustrated, Fearful, Humili-ated, and Upset, with correlation coefficients of 0.43, 0.5,0.53, 0.55, and 0.61, respectively, and which aremarked with blue blocks. The values of all correlationcoefficients are greater than 0, which indicate positivecorrelations—that is to say, the corresponding twoemotions will strengthen or weaken at the same time.

Fig. 6 A two-way ANOVA thermogram of age and embarrassment components

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 9 of 15

In view of the analysis results of Fig. 9, the traditionalmultivariate linear regression cannot be used for an ac-curate calculation, due to the multicollinearity betweenthe independent variables of the emotions. Therefore,this study adopted the regression algorithm of LASSO(least absolute shrinkage and selection operator), whichcan improve the interpretability and accuracy of themodel (Song et al., 2016). The regression equation ofEmbarrassment is

Embarrassment ¼ 0:687Shameþ 0:108Anxiousþ 0:226Upset

Here, the best alpha is 0.081, and R2 is 0.758.

DiscussionThis study takes sent WeChat emoji with opposite mean-ings as an example for discussing the embarrassment of thesender. “Embarrassment” is a situation that cannot be

avoided in interpersonal relationships; it reflects a person’sfeeling of the appropriateness of behavior, causes a person’sfrustration, and forms a negative emotion (Babcock, 1988).In this study, the degrees of explanation of shame for De-pressed, Anxious, Upset, Shocked, Foolish, and Humiliatedfor Embarrassment are all above 0.82, which is high.Shame has the highest explanation for Embarrassment,

which is consistent with the opinion of Borg et al.(1988). Embarrassment is always considered as a dimen-sion of Shame (Oleszkiewicz et al., 2017). In other lan-guages of modern society, Embarrassment is regarded asa minor version of Shame. For example, “verguenza” inSpanish has two meanings: embarrassment and shame.In many languages, the commonly used word “shame” isalso a kind of embarrassment (Evans, 2015).Scholars have pointed out that the anxiety is an import-

ant feature of embarrassment (Jankowski & Takahashi,2014). Among general people, higher levels of anxiety areassociated with greater embarrassment (Bas-Hoogendam

Fig. 7 The mean value of respondents of three age groups for different emotions under 10 dialogue scenarios

Fig. 8 Relationship between embarrassment and emotions

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 10 of 15

et al., 2018), and social anxiety is closely related to embar-rassment (Hofmann et al., 2006). Previous research(Miller, 1996) has defined embarrassment more clearly asa self-perceptible emotion that has a blurring line withanxiety and shame.“Embarrassment” makes people feel constrained, which

takes a period of time to recover from. Embarrassmentalso destroys a pleasant mood. Once embarrassment oc-curs, people enter a state of panic, clumsiness, shame, andchagrin, which makes them feel very uncomfortable(Miller, 1996). Fifty percent of the people with embar-rassed feelings will also feel humiliation (Grace, 2007).When users mistakenly send WeChat emojis with an op-

posite meaning, it can cause embarrassment and have anegative impact on relationships, but sometimes it cannotbe said to be harmful. It is an emotional response to thesudden occurrence of seemingly harmless, and sometimeshumorous, events (Miller, 1996). Embarrassment makes theindividual feel stupid (Goffman, 1967).The Embarrassment of mistakenly sending an emoji with

opposite meanings involves self-esteem. People with differentdegrees of self-esteem react differently to embarrassment(Song et al., 2017), which is caused by the potential loss ofself-esteem (Modigliani, 1971), including the negative evalu-ation of others (Modigliani, 1971), one’s own behavior notconforming to certain ideal traits (Babcock, 1988), the failureto present a consistent image (Parrott & Smith, 1991), a viola-tion of other people’s expectations, etc. (Borg et al., 1988).“Embarrassment” makes people feel constrained, which

takes a period of time to recover from. Embarrassment also

destroys a pleasant mood. Once embarrassment occurs,people enter a state of panic, clumsiness, shame, and chag-rin, which makes them feel very uncomfortable (Miller,1996). This study explains the embarrassing psychologicalprocesses and emotional components of the WeChat emojiwith opposite meanings that were mistakenly sent.People usually want to leave a good impression in their

social circle, but when they make careless mistakes whileattending activities, they will feel embarrassed andashamed (Kaufman, 2004; Lewis, 1971), as was the caseof the emotional state of the sender in this study aftersending WeChat emoji incorrectly.These findings highlighted that the key components

are Shame, Depressed, Anxious, Upset, Shocked, Foolish,and Humiliated, which deserve enough attention andplay an important role in improving the users’ willing-ness to continue using WeChat. In particular, reducingthe users’ emotional reactions of shame helps to improvethe users’ willingness to continue using WeChat.Compared with the other embarrassment of compo-

nents discussed above, namely, Self-conscious, Frustrated,Fearful, Distressed and Anger, this explains that the de-gree of embarrassment is low. According to previous re-search on embarrassment that is related to consumption,anger accounts for the lowest proportion of embarrassingemotions. In studies on the embarrassing emotional reac-tions caused by missending emojis, as compared withother emotions, anger can less likely explain the change ofembarrassment. Research has produced different results,which represents the complexity of emotions (Borg et al.,

Fig. 9 Thermodynamic diagram of correlation coefficients

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 11 of 15

1988). It also shows that in WeChat’s emoji, which mis-takenly sends the opposite meaning, anger can explainembarrassment, but does not bring a strong angry embar-rassment to the sender.In this study, although gender and age had no signifi-

cant influence on embarrassment, a further analysisshowed that people of different genders and ages wereaffected to different degrees by embarrassment.When sending WeChat emoji with opposite meanings,

males are more easily affected by embarrassment, com-pared with females. This can be explained by the fact thatalthough females use more emojis than males (Wu et al.,2019), they are stronger than men in enduring certain emo-tional stress. Males will be significantly affected by psycho-logical factors, while females will not, and their strengthcontinues to grow (Lin & Yin, 2015). The results of thisstudy show that men were more embarrassed than womenwhen they mistakenly sent an emoji with an oppositemeaning. However, most previous studies have confirmedthat women are more embarrassed than men (Costa et al.,2001; Hall, 2011; Miller, 1992; Miller, 1987). The findingsof this study are novel and have some implications for fur-ther theoretical research and the development of WeChatand other social networking services. According to the re-search results, 18–25-year-old users are more susceptible toembarrassment when missending WeChat emojis, com-pared with 31–40-year-old users. Previous research haspointed out that young people, aged between 18 and 25years, often use emojis and say that these can express theiremotions better (Zhang et al., 2017). As for the number ofparticipants in this study, the majority of participants werebetween 18 and 25 years old. The use of emojis decreaseswith age (Dhir & Tsai, 2017; Ozimek et al., 2017; Zhanet al., 2016). There is a correlation between emojis with op-posite meanings and the frequency of emoji usage. The useof emojis is related to age (Wang et al., 2018), as youngergenerations (i.e., college students) are very active users ofMIMs (Oleszkiewicz et al., 2017). WeChat is the main mo-bile and MIM application among young Chinese people(Cohen-Zada & Krumer, 2017). People in the 18–25-year-old age group were also more likely to be embarrassed bymistakenly sending emoji with opposite meanings. Thesefindings will also provide a reference for MIM developers.Our research, like other studies, has its research limita-

tions. First, it is difficult to collect samples that meet the re-quirements of this study. Although the number of samplesmeets the planned number of studies, the number of sam-ples is still small, compared with other studies. It is import-ant to acknowledge that our sample was small, generatingsmaller power in our results (Laurence et al., 2020). Sec-ondly, when we define emoji with an opposite meaning, wemainly consider a specific meaning, context, and languageculture. However, there may be other factors that affect peo-ple’s understanding of emoji with opposite meanings.

Thirdly, as this study takes WeChat as an example, its scopeis limited to a single language and culture background andindependent instant messaging applications. Therefore, wecannot understand the difference of embarrassment causedby sending emoji with opposite meanings in other lan-guages, cultures, and instant messaging software.

Conclusion and suggestionsBecause embarrassment has adverse consequences, elimin-ating it in WeChat communication can enhance the inter-action between users and make them willing to continue touse the social platform. As previous scholars have studiedone or several important embarrassment-related emotions,this study explored the embarrassment caused by missend-ing emojis with opposite meanings, and it has provided sev-eral important embarrassment-related emotions.Surprisingly, we found that males in the study were moresusceptible to embarrassment than females. In addition,when peers use WeChat, people are more motivated tomaintain relationships (Borg et al., 1988; Kaufman, 2004;Lewis, 1971). It should be further verified in future studieswhether it is possible that males are more willing to main-tain social relationships than females in WeChat informa-tion interaction. Young users aged between 18 and 30 yearsare more susceptible to embarrassment when they mis-takenly send WeChat emojis. On missending WeChatemojis with opposite meanings, the major components thatexplain Embarrassment are Shame, Anxious, and Upset.Reducing the users’ emotional reaction of shame can im-prove their willingness to continuously use WeChat. Wethus encourage real-time communication app developers tomake full use of these findings, to design the right emojialignment to reduce user embarrassment, to promote thefriendly social interaction of real-time communication apps,and to achieve the good sustainable development of net-work social interaction.According to the research limitations, we propose, firstly,

that future research should collect more samples that meetthe research requirements by using effective samplecollection methods. Secondly, when defining the op-posite meanings of emoji, researchers should considerthe influence of other aspects. For example, themeaning of emoji may be related to social geography(Scheff, 2015). Considering the influence of more fac-tors on the opposite meaning of emoji will help togeneralize the research results. Thirdly, in the futureresearch, we should compare the differences and con-nections of the opposite meanings of emoji in differ-ent languages and cultural backgrounds. In addition,we should analyze the differences between the users’understanding of the opposite meaning of emoji indifferent instant messaging software. Based on theabove two aspects, researchers can compare differentresearch results to obtain new conclusions.

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 12 of 15

AcknowledgementsOur deepest gratitude goes to the anonymous reviewers for their carefulreview and constructive comments, which have greatly improved this paper.We are also very grateful to the editor for his careful work and thoughtfuladvice.

Authors’ contributionsLL and QJ conceived and planned the experiments. LL and QJ wereresponsible for data collection. LL and ZJ conducted data analyses andcontributed to the interpretation of the results. ZJ performed datavisualization processing. LL took the lead in writing the manuscript. YHcollaborated in the writing of the paper and the editing of the finalmanuscript. YH also performed critical revision of the paper. The authorsread and approved the final manuscript.

FundingThis work was supported by the Zhejiang Social Sciences under Grant16NDJC106YB.

Availability of data and materialsThe datasets used and/or analyzed during the current study are availablefrom the corresponding author on reasonable request.

Competing interestsThe authors declare that they have no competing interests.

Author details1College of Science & Technology, Ningbo University, Ningbo, China. 2TheGraduate Institute of Design Science, Tatung University, Taipei, Taiwan.3School of Design, The Hong Kong Polytechnic University, Hong Kong,China. 4School of Design, Jiangnan University, Wuxi, China.

Received: 30 April 2020 Accepted: 6 August 2020

ReferencesAnnamalai S & Salam SN. (2017). Undergraduates' interpretation on WhatsApp

smiley emoji. Jurnal Komunikasi, Malaysian Journal of Communication, 33(4),89-103. https://doi.org/10.17576/JKMJC-2017-3304-06

Babcock, MK. (1988). Embarrassment: a window on the self. Journal for the Theoryof Social Behaviour, 18(4), 459-483. https://doi.org/10.1111/j.1468-5914.1988.tb00510.x

Bas-Hoogendam, JM, van Steenbergen, H, van der Wee, NJA, & Westenberg, PM.(2018). Not intended, still embarrassed: social anxiety is related to increasedlevels of embarrassment in response to unintentional social norm violations.European Psychiatry, 52, 15-21. https://doi.org/10.1016/j.eurpsy.2018.03.002

Bastin, C, Harrison, BJ, Davey, CG, Moll, J, & Whittle, S. (2016). Feelings of shame,embarrassment and guilt and their neural correlates: a systematic review.Neuroscience & Biobehavioral Reviews, 71, 455-471. https://doi.org/10.1016/j.neubiorev.2016.09.019

Bethell, L, Lin, HC, & McFatter, R. (2014). Embarrassment and empathy beforehelping: how internal working models come into play. Motivation and Emotion,38(1), 131-139. https://doi.org/10.1007/s11031-013-9369-y

Borg, I., Staufenbiel, T., & Scherer, K. R. (1988). On the symbolic basis of shame. InFacets of emotion: Recent research, (pp. 79–98). Hillsdale, NJ, US: LawrenceErlbaum Associates, Inc..

Castillo, C, Mendoza, M, & Poblete, B. (2013). Predicting information credibility intime-sensitive social media. Internet Research, 23(5), 560-588. https://doi.org/10.1108/IntR-05-2012-0095

Chen, Z, Lu, X, Shen, S, Ai, W, Liu, X, & Mei, Q. (2017). Through a gender lens: anempirical study of emoji usage over large-scale android users. Retrieved fromhttps://arxiv.org/abs/1705.05546

Choi, YK, Seo, Y, & Yoon, S. (2017). E-WOM messaging on social media: social ties,temporal distance, and message concreteness. Internet Research, 27(3), 495-505. https://doi.org/10.1108/IntR-07-2016-0198

Cohen-Zada, D, & Krumer, A. (2017). Psychological momentum and gender.Journal of Economic Behavior & Organization, 135, 66-81. https://doi.org/10.1016/j.jebo.2017.01.009

AppendixTable 5 Ten dialogue scenarios and explanations

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 13 of 15

Costa, M, Dinsbach, W, Manstead, ASR., & Bitti, PER. (2001). Social presence,embarrassment, and nonverbal behavior. Journal of Nonverbal Behavior, 25(4),225-240. https://doi.org/10.1023/A:1012544204986

Danesi, M. (2016). The semiotics of emoji: The rise of visual language in the age ofthe internet. London: Bloomsbury Publishing.

Darics, E. (2013). Non-verbal signalling in digital discourse: the case of letterrepetition. Discourse, context & media, 2(3), 141-148. https://doi.org/10.1016/j.dcm.2013.07.002

Derks, D, Bos, AER, & Von Grumbkow J. (2007). Emoticons and social interactionon the Internet: the importance of social context. Computers in HumanBehavior, 23(1), 842-849. https://doi.org/10.1016/j.chb.2004.11.013

Dhir, A, & Tsai, CC. (2017). Understanding the relationship between intensity andgratifications of Facebook use among adolescents and young adults.Telematics and Informatics, 34(4), 350-364. https://doi.org/10.1016/j.tele.2016.08.017

Evans, V. (2015). No, the rise of the emoji doesn’t spell the end of language. TheConversation. https://theconversation.com/no-the-rise-of-the-emoji-doesnt-spell-the-end-of-language-42208

Freedman, A. (2018). Cultural literacy in the empire of emoji signs: Who is cryingwith joy? First Monday, 23. https://doi.org/10.5210/fm.v23i9.9395

Ganster, T, Eimler, SC., & Krämer, NC. (2012). Same same but different!? Thedifferential influence of smilies and emoticons on person perception.Cyberpsychology, Behavior, and Social Networking, 15(4), 226-230. https://doi.org/10.1089/cyber.2011.0179

Goffman, E. (1967). Interaction ritual: Essays in face-to-face behavior. New York:Anchor Books.

Grace, D. (2007). How embarrassing! An exploratory study of critical incidentsincluding affective reactions. Journal of Service Research, 9(3), 271-284. https://doi.org/10.1177/109467050700900305

Hall, JA. (2011). Is it something I said? Sense of humor and partnerembarrassment. Journal of Social and Personal Relationships, 28(3), 383-405.https://doi.org/10.1177/0265407510384422

Hanna, E, Ward, L M, Seabrook, RC, Jerald, M, Reed, L, Giaccardi, S, & Lippman, JR.(2017). Contributions of social comparison and self-objectification inmediating associations between Facebook use and emergent adults'psychological well-being. Cyberpsychology, Behavior, and Social Networking,20(3), 172-179. https://doi.org/10.1089/cyber.2016.0247

Higuchi, M, & Fukada, H. (2002). A comparison of four causal factors ofembarrassment in public and private situations. The Journal of Psychology,136(4), 399-406. https://doi.org/10.1080/00223980209604166

Hofmann, SG, Moscovitch, DA, & Kim, Hyo-Jin. (2006). Autonomic correlates ofsocial anxiety and embarrassment in shy and non-shy individuals.International journal of Psychophysiology, 61(2), 134-142. https://doi.org/10.1016/j.ijpsycho.2005.09.003

Hongqiang, S, Xichen, L, Ke, L, & Jingzhong, X. (2019, Dec 17-19). Emoji usageand interpersonal relationship in computer-mediated communication. 2019International Joint Conference on Information, Media and Engineering(IJCIME), Osaka, Japan.

Hossain, Md, Jahan, N, Fang, Y, & Hoque, S. (2019). Nexus of electronic word-of-mouthto social networking sites: a sustainable chatter of new digital social media.Sustainability, 11(3), 759. https://doi.org/10.3390/su11030759

Hu, T, Guo, H, Sun, H, Nguyen, TT, & Luo, J. (2017, May 15-18). Spice up your chat:the intentions and sentiment effects of using emojis. Eleventh InternationalAAAI Conference on Web and Social Media, Montreal, Quebec, Canada.

Jaeger, SR, Xia, Y, Lee, PY, Hunter, DC, Beresford, MK, & Ares, G. (2018). Emojiquestionnaires can be used with a range of population segments: findings relatingto age, gender and frequency of emoji/emoticon use. Food Quality and Preference,68, 397-410. https://doi.org/10.1016/j.foodqual.2017.12.011

Jankowski, KF, & Takahashi, H. (2014). Cognitive neuroscience of social emotionsand implications for psychopathology: examining embarrassment, guilt, envy,and schadenfreude. Psychiatry and Clinical Neurosciences, 68(5), 319-336.https://doi.org/10.1111/pcn.12182

Kaufman, G. (2004). The psychology of shame: Theory and treatment of shame-based syndromes. New York: Springer.

Kaye, LK., Wall, HJ., & Malone, SA. (2016). “Turn that frown upside-down”: acontextual account of emoticon usage on different virtual platforms.Computers in Human Behavior, 60, 463-467. https://doi.org/10.1016/j.chb.2016.02.088

Kelly, R., & Watts, L. (2015). Characterising the inventive appropriation of emoji asrelationally meaningful in mediated close personal relationships. In

Experiences of technology appropriation: Unanticipated users, usage,circumstances, and design. Oslo: Norway.

Laurence PG, Busin Y, da Cunha Lima HS & Macedo EC. (2020). Predictors ofproblematic smartphone use among university students. Psicologia: Reflexãoe Crítica, 33(1), 8. https://doi.org/10.1186/s41155-020-00147-8

Lewis, H. B. (1971). Shame and guilt in neurosis. Psychoanalytic review, 58(3), 419–438.

Li, L, & Yang, Y. (2018). Pragmatic functions of emoji in internet-basedcommunication---a corpus-based study. Asian-Pacific Journal of Second andForeign Language Education, 3(1), 16. https://doi.org/10.1186/s40862-018-0057-z

Lin, F. (2019). Positive or negative: emoji usage in online social media. 4thInternational Conference on Humanities Science, Management andEducation Technology (HSMET 2019), Singapore.

Lin, R, & Yin, G. (2015). Bayes factor and posterior probability: complementarystatistical evidence to p-value. Contemporary clinical trials, 44, 33-35. https://doi.org/10.1016/j.cct.2015.07.001

Ljubešić, N., & Fišer, D. (2016). A global analysis of emoji usage. In Proceedings ofthe 10th Web as Corpus Workshop. Germany: Berlin.

Lomanowska, AM, & Guitton, MJ. (2016). Online intimacy and well-being in thedigital age. Internet Interventions, 4, 138-144. https://doi.org/10.1016/j.invent.2016.06.005

McCambridge, SA, & Consedine, NS. (2014). For whom the bell tolls:Experimentally-manipulated disgust and embarrassment may causeanticipated sexual healthcare avoidance among some people. Emotion, 14(2),407. https://doi.org/10.1037/a0035209

Miguel, FK, & Pessotto, F. (2016). Projective aspects on cognitive performance:distortions in emotional perception correlate with personality. Psicologia:Reflexão e Crítica, 29(1), 17. https://doi.org/10.1186/s41155-016-0036-6

Miller, RS. (1987). Empathic embarrassment: Situational and personaldeterminants of reactions to the embarrassment of another. Journal ofPersonality and social Psychology, 53(6), 1061-1069. https://doi.org/10.1037/0022-3514.53.6.1061

Miller, RS. (1992). The nature and severity of self-reported embarrassingcircumstances. Personality and Social Psychology Bulletin, 18(2), 190-198.https://doi.org/10.1177/0146167292182010

Miller, R. S. (1996). Embarrassment: Poise and peril in everyday life. Guilford Press.Modigliani, A. (1971). Embarrassment, facework, and eye contact: testing a theory

of embarrassment. Journal of Personality and social Psychology, 17(1), 15.https://doi.org/10.1037/h0030460

Montag, C, Becker, B, & Gan, C. (2018). The multipurpose application WeChat: areview on recent research. Frontiers in psychology, 9, 2247. https://doi.org/10.3389/fpsyg.2018.02247

Morris, D. (1979). Gestures, their origins and distribution. New York, NY: Stein & DayPub.

Oleszkiewicz, A, Karwowski, M, Pisanski, K, Sorokowski, P, Sobrado, B, &Sorokowska, A. (2017). Who uses emoticons? Data from 86 702 Facebookusers. Personality and Individual Differences, 119, 289-295. https://doi.org/10.1016/j.paid.2017.07.034

Ozimek, P, Baer, F, & Förster, J. (2017). Materialists on Facebook: the self-regulatory role of social comparisons and the objectification of Facebookfriends. Heliyon, 3(11), e00449. https://doi.org/10.1016/j.heliyon.2017.e00449

Pang, H. (2016). Understanding key factors affecting young people's WeChatusage: an empirical study from uses and gratifications perspective.International Journal of Web Based Communities, 12(3), 262-278. https://doi.org/10.1504/IJWBC.2016.077757

Pang, H. (2018a). How does time spent on WeChat bolster subjective well-beingthrough social integration and social capital? Telematics and Informatics,35(8), 2147-2156. https://doi.org/10.1016/j.tele.2018.07.015

Pang, H. (2018b). Understanding the effects of WeChat on perceived socialcapital and psychological well-being among Chinese international collegestudents in Germany. Aslib Journal of Information Management, 70(3), 288-304. https://doi.org/10.1108/AJIM-01-2018-0003

Parrott, WG, & Smith, SF (1991). Embarrassment: actual vs. typical cases, classicalvs. prototypical representations. Cognition and Emotion, 5(5-6), 467-488.https://doi.org/10.1080/02699939108411053

Pauleen, David J, & Yoong, Pak. (2001). Facilitating virtual team relationships viaInternet and conventional communication channels. Internet Research, 11(3),190-202. https://doi.org/10.1108/10662240110396450

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 14 of 15

Pavalanathan, U, & Eisenstein, J. (2016). More emojis, less :) the competition forparalinguistic function in microblog writing. First Monday, 21(11). https://doi.org/10.5210/fm.v21i11.6879

Research, I. (2016). China mobile social communications research reports (2015Q3).I. Research. http://www.iresearch.com.cn/report/2520.html. Accessed Feb2019.

Riordan, MA (2017a). The communicative role of non-face emojis: affect anddisambiguation. Computers in Human Behavior, 76, 75-86. https://doi.org/10.1016/j.chb.2017.07.009

Riordan, MA. (2017b). Emojis as tools for emotion work: communicating affect intext messages. Journal of Language and Social Psychology, 36(5), 549-567.https://doi.org/10.1177/0261927X17704238

Scheff, T. (2015). Toward defining basic emotions. Qualitative Inquiry, 21(2), 111-121. https://doi.org/10.1177/1077800414550462

Sharkey, W, & Singelis, T (1995). Embarrassability and self-construal: a theoreticalintegration. Personality and Individual Differences, 19(6), 919-926. https://doi.org/10.1016/S0191-8869(95)00125-5

Song, X, Huang, F, & Li, X. (2017). The effect of embarrassment on preferences forbrand conspicuousness: the roles of self-esteem and self-brand connection.Journal of Consumer Psychology, 27(1), 69-83. https://doi.org/10.1016/j.jcps.2016.05.001

Song, Y, Dai, XY, & Wang, J. (2016). Not all emotions are created equal: expressivebehavior of the networked public on China's social media site. Computers inHuman Behavior, 60, 525-533. https://doi.org/10.1016/j.chb.2016.02.086

Tannen, D. (1994). Gender and discourse. England: Oxford University Press.Tigwell, G. W., & Flatla, D. R. (2016, September). Oh that's what you meant!:

reducing emoji misunderstanding. In Proceedings of the 18th internationalconference on human-computer interaction with mobile devices and servicesadjunct. Florence: Italy.

Tsai, WHS, & Men, RL. (2018). Social messengers as the new frontier oforganization-public engagement: a WeChat study. Public Relations Review,44(3), 419-429. https://doi.org/10.1016/j.pubrev.2018.04.004

Tseng, TH, & Hsieh, SH. (2019). Determinants of emoticon usage in mobile instantmessaging: a construal level theory perspective. Behaviour & InformationTechnology, 38(3), 289-301. https://doi.org/10.1080/0144929X.2018.1531927

Uysal, A, Akbaş, G, Helvacı, E, & Metin, İ. (2014). Validation and correlates of thevicarious embarrassment scale. Personality and Individual Differences, 60, 48-53. https://doi.org/10.1016/j.paid.2013.12.015

Wang, G, Zhang, W, & Zeng, R. (2019). WeChat use intensity and social support:the moderating effect of motivators for WeChat use. Computers in HumanBehavior, 91, 244-251. https://doi.org/10.1016/j.chb.2018.10.010

Wang, S, Ji, B, Zhao, J, Liu, W, & Xu, T. (2018). Predicting ship fuel consumptionbased on LASSO regression. Transportation Research Part D: Transport andEnvironment, 65, 817-824. https://doi.org/10.1016/j.trd.2017.09.014

Wen, Z, Geng, X, & Ye, Y. (2016). Does the use of WeChat lead to subjective well-being? The effect of use intensity and motivations. Cyberpsychology, Behavior,and Social Networking, 19(10), 587-592. https://doi.org/10.1089/cyber.2016.0154

Wu, J, Liao, H, Wang, JW, & Chen, Tianqi. (2019). The role of environmentalconcern in the public acceptance of autonomous electric vehicles: a surveyfrom China. Transportation Research Part F: Traffic Psychology and Behaviour,60, 37-46. https://doi.org/10.1016/j.trf.2018.09.029

Yang, C, & Lee, Y. (2020). Interactants and activities on Facebook, Instagram, andTwitter: associations between social media use and social adjustment tocollege. Applied Developmental Science, 24(1), 62-78. https://doi.org/10.1080/10888691.2018.1440233

Zhan, L, Sun, Y, Wang, N, & Zhang, X. (2016). Understanding the influence ofsocial media on people’s life satisfaction through two competingexplanatory mechanisms. Aslib Journal of Information Management, 68(3),347-361. https://doi.org/10.1108/AJIM-12-2015-0195

Zhang, CB, Li, YN, Wu, B, & Li, DJ. (2017). How WeChat can retain users: roles ofnetwork externalities, social interaction ties, and perceived values in buildingcontinuance intention. Computers in Human Behavior, 69, 284-293. https://doi.org/10.1016/j.chb.2016.11.069

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Liu et al. Psicologia: Reflexão e Crítica (2020) 33:20 Page 15 of 15